为了解决并联机器人"盲"的工作状态,构建了球形仿生眼监测平台.基于球形仿生眼监测平台给出一种并联机器人加工刀具匹配方法:通过图像预处理,获得刀具图像的边缘信息,并将边缘线拐点作为特征点;根据极线约束原理,对特征点进行粗粒度匹配,获取准同名点;提取以准同名点为中心的短边缘线段并对其进行区域校正,通过细粒度匹配,对同名点进行准确定位.采用该方法对并联机器人加工刀具实现了特征点的匹配.实验表明:图像匹配的速度和精度有了较大提高,刀具加工点的三维重建比较精确.
To solve the blind working state of the parallel robot,the spherical biomimetic eye monitoring platform is proposed.Based on the spherical biomimetic eye monitoring platform,a parallel robot tool match method is presented and implemented.The tool edge images could be gotten by pretreatment.Meanwhile,the inflection points of the edge line are regarded as the feature points.According to stereo vision theory,the quasi homonymy point could be acquired by the coarse-grained matching.Then the shorter edge line whose center is the quasi homonymy point is acquired.And the region which shorter edge line in is correcet.The homonymy points are positioned by the fine-grained matching.By this means,the characteristic point match is completed.The experimental results show that the matching speed and accuracy are improved especially,and the 3D reconstruction of the tool work point become more accurate.